linux-kernel - Re: [PATCH] remove volatile from nmi.c

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Fri, 14 Jul 2006 23:28:29 +1000
From:	Nick Piggin <nickpiggin@...oo.com.au>
To:	Steven Rostedt <rostedt@...dmis.org>
CC:	Andrew Morton <akpm@...l.org>, Ingo Molnar <mingo@...e.hu>,
	Linus Torvalds <torvalds@...l.org>,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] remove volatile from nmi.c

Steven Rostedt wrote:
> OK, I'm using this as something of an exercise to completely understand
> memory barriers.  So if something is incorrect, please let me know.
> 
> This patch removes the volatile keyword from arch/i386/kernel/nmi.c.
> 
> The first removal is trivial, since the barrier in the while loop makes
> it unnecessary. (as proved in "[patch] spinlocks: remove 'volatile'"
> thread)
> http://marc.theaimsgroup.com/?l=linux-kernel&m=115217423929806&w=2
> 
> 
> The second is what I think is correct.  So please review.

Please comment memory barriers if possible.

The second adds memory barriers that weren't there before... how come
they are needed? Basically, the comment should be a pointer to the
read side, or illustrate a typical readside where write reordering
would cause a problem.

> 
> Thanks,
> 
> -- Steve
> 
> Signed-off-by: Steven Rostedt <rostedt@...dmis.org>
> 
> Index: linux-2.6.18-rc1/arch/i386/kernel/nmi.c
> ===================================================================
> --- linux-2.6.18-rc1.orig/arch/i386/kernel/nmi.c	2006-07-14 08:35:00.000000000 -0400
> +++ linux-2.6.18-rc1/arch/i386/kernel/nmi.c	2006-07-14 08:38:07.000000000 -0400
> @@ -106,7 +106,7 @@ int nmi_active;
>   */
>  static __init void nmi_cpu_busy(void *data)
>  {
> -	volatile int *endflag = data;
> +	int *endflag = data;
>  	local_irq_enable_in_hardirq();
>  	/* Intentionally don't use cpu_relax here. This is
>  	   to make sure that the performance counter really ticks,
> @@ -121,7 +121,7 @@ static __init void nmi_cpu_busy(void *da
>  
>  static int __init check_nmi_watchdog(void)
>  {
> -	volatile int endflag = 0;
> +	int endflag = 0;
>  	unsigned int *prev_nmi_count;
>  	int cpu;
>  
> @@ -150,7 +150,7 @@ static int __init check_nmi_watchdog(voi
>  			continue;
>  #endif
>  		if (nmi_count(cpu) - prev_nmi_count[cpu] <= 5) {
> -			endflag = 1;
> +			set_wmb(endflag, 1);
>  			printk("CPU#%d: NMI appears to be stuck (%d->%d)!\n",
>  				cpu,
>  				prev_nmi_count[cpu],
> @@ -161,7 +161,7 @@ static int __init check_nmi_watchdog(voi
>  			return -1;
>  		}
>  	}
> -	endflag = 1;
> +	set_wmb(endflag, 1);
>  	printk("OK.\n");
>  
>  	/* now that we know it works we can reduce NMI frequency to


-- 
SUSE Labs, Novell Inc.
Send instant messages to your online friends http://au.messenger.yahoo.com 
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/